Re: [-empyre-] Writing and Pattern Flows

[John Klima <john_klima@cityarts.com>]

gee whiz... wow.

Bill makes a number of points (if I understand them correctly) that 
directly (or in some cases, conversely) relate to my own recent applied 
studies. I'm gonna take the liberty to riff a little here, I'm sure he 
wont mind. I have had very similar things on my mind as of late,  though 
as they apply to "real world," or more accurately "game world" problems.

" The notion of fields of meaning is central to meaning becoming, where 
each-media element functions as a field of potential meaning, having a 
meaning force..."

In autonomous agent technology, a common steering paradigm is known as 
"flow field following," where a "surface" contains regions of 
attraction/repulsion, and bots are set-off to traverse through this 
"landscape."  I use a physical metaphor as a means of visualization, but 
these regions and their "force fields" can be visible/physical, or not.  
A similar metaphor might be Einstein's theory of gravity - depressions 
in an otherwise flat fabric of continuum - "meaning becoming" being our 
bots, or celestial bodies in the universe.

The potential exists to apply this notion of flow field to a dynamic 
narrative trajectory, where a history has an initial vector,  is run 
through a field of narrative possibilities,  and the field of forces 
result in an output vector. Specific possibilities/states can be given 
proportional weights that sum to produce the output vector. These 
weights are fungable as time and other forces modify the landscape 
(including the initial trajectory itself - everything has a wake).  I'm 
currently actively exploring this potential for dynamic narrative 
generation.

As Bill suggests, this weighted flow field could be applied to form an 
agent's domain knowledge, again in terms of an input vector, the fields  
(which the bot can never have complete awareness of, a bot is not god) 
create a specific output vector of domain awareness. The vector is a 
specific percept/conclusion of a "too huge to be completely knowable" 
world. given our current mechanical capabilities, impressive though they 
may be, that "too huge" is a shockingly small quantity.  Again, i'm 
actively engaged in constructing such architectures as a means of 
producing rational behavior in bots.

"Computers are particularly good at generating pattern flows and thus 
fall within the sphere of meaning production...."

If it can generate a flow pattern, it can conversely consume a flow 
patern (though this tends to be a far more difficult task). This is, in 
effect, exactly what a neural network does. Contained in the hidden 
layer of a hopfield network, exists a topology of weights that recieve 
an input node, and accumulate "force" until a certain threshold value 
causes the hidden nodes to fire to an output layer (which can then be 
fed as the inputs to yet another network). the force field is fungable, 
and through back-propagation can be "trained" and additionally, 
"retrained." The network, in a constant state of "meaning becoming" 
though not possessing an absolute awareness of it's domain (ie its not 
god), can nonetheless draw some pretty reasonable conclusions from its 
percepts.

best,
j




Bill Seaman wrote:

> Bill Seaman
>
> The central concern of my work since 1980 has been meta-meaning 
> production. In particular I am interested in the functionality of 
> language in our coming to know the world. I have developed a number of 
> generative systems exploring the production of meaning. In terms of 
> "writing" I have been using the term "pattern flows" to refer to and 
> reflect on embodied knowing. I am very much interested in how 
> multi-modal flows of experience - pattern flows - inform our meaning 
> becoming. I understand mind and body to be co-arising with the 
> functionality of pattern flows (an extended linguistics) as being 
> central to this process. Computers are particularly good at generating 
> pattern flows and thus fall within the sphere of meaning production. 
> In fact computers enable the continuous growth of our relation to 
> linguistic construction. The projective employment of language is 
> central to ongoing experience. The notion of fields of meaning is 
> central to meaning becoming, where each-media element functions as a 
> field of potential meaning, having a meaning force. These pattern flow 
> forces are summed by the body in an ongoing manner. Each media-element 
> has attributes which are "of themself". Thus the computer as a pattern 
> production mechanism enables one to juxtapose pattern-flows with many 
> qualities simultaneously - words, sounds, music, images, 3D models, 
> illustrations, diagrams, time-based video recordings etc. etc. This 
> meaning production sphere is expanding and beginning to include the 
> haptic and other physical potentialities. Writing cannot be separated 
> out from either pattern flow acquisition or multi-modal pattern 
> production. Writing always becomes enfolded in meaning production as a 
> living associational relation informed by the history of multi-modal 
> experience as it intermingles with current context. Thus words carry a 
> meaning force that co-arises with ongoing experience.
>
> I think of digital writing as employing all manner of digital pattern 
> flows in the service of evocation. I also see the potential of 
> exploring the writing space that bridges the continuum between the 
> physical and the digital as an exciting field of inquiry. Each media 
> element be they digital or other, informing the ongoing meaning summing.
>
>
>
> Interface is a very vague term...